The SolidSim Project - CO-LaN– DIVA, gProms, SpeedUp, Aspen Custom Modeler –sequential modular...

CO-LaN Meeting in Leipzig-Schkopau

Verfahrenstechnik I Verfahrenstechnik III

The SolidSim Project

Development of a Flowsheeting Tool

for Solids Processes

CO-LaN Meeting 13.02.2004


1

Outline

• The SolidSim Project

– Project introduction

– Design concept of Simulation Software

• CAPE-OPEN and SolidSim

• Demonstration of SolidSim



2

Motivation

• Flowsheet simulator for solids processes allows

– development of flowsheets for complex processes

– design studies for process alternatives

– sensitivity analysis for process parameters

– fast and easy optimization of processes involving solids

• Establish a model library for solids processing

equipment and unit operations

• Support education of chemical engineering students



3

Comparision - Simulation of …

• Liquid/gas process– concentrated variables

(temp., pressure, mass fraction)

– complete characterization by few

parameters(temp., composition, pressure)

– models cover mass transfer and

reactions

– equipment geometry plays minor

role in modeling (just influence on residence time)

– throughput influences only

residence time

• Solids processing– additional distributed parameters

(particle size distr., density distr.)

– characterization requires more

parameters(shape factor, size distribution,

porosity, breaking behaviour, etc.)

– partially dependant secondary

attributes(particle size dep. composition (ore)

or moisture content, surface area

dependant contamination)

– model has to account for very

complex physical effects(e.g. impact mill, dryer, fluidized bed)

• most models only valid for certain

operating conditions

Solids require a very complex

information structure



4

Commercial Simulators I

• Commercial simulators can be divided into two

groups:

– equation based systems

• system is modeled as one large equation system and solved

parallel for all variables

– DIVA, gProms, SpeedUp, Aspen Custom Modeler

– sequential modular aproach

• process is divided into connected logical blocks as in a flowsheet

• blocks are then solved individually

• recycles solved by iterative solution of the system

– ASPEN Plus, Hyprotech HySim



5

Commercial Simulators II

• Problems of available sequential modular systems:

– Model quality/quantity

• generally only a few models available for solids processing

• models usually have low detail level (short-cut models, empirical

correlations)

– Stream structures supported

• most simulators originate from liquid/gas process industry

• concentrated parameters such as temperature, pressure, etc. are

supported, plus in some cases simple support for distributed

parameters such as PSD, but no secondary dependend attributes



6

Commercial Simulators III

• Characteristics of equation based simulators

– High flexibility due to equation based design:

• almost any model can be implemented as long as equations are

given and solver supports type of equations

• custom parameters can be introduced with ease

– Input of problem requires knowledge of “programming

language” → syntax, conventions, etc.

– Every single model has to be implemented by including

the equations → high effort for a simple problem

– Equation sets get very large, flowsheeting based

simulator is closer to engineers view on process



7

Requirements for a Simulator

• support of flexible information structure for streams (distributed and concentrated parameters, standard and custom parameters)

• allow combination of models for individual equipment and process steps to be combined to complex process flowsheet, no limitation to special processes

• presence of fluids and solid/liquid interactions have to be considered

• ergonomic easy-to-use Graphical User Interface (GUI) to allow engineer to build process following his view (flowsheet focussed)

• extendable model database including model backgrounds (mathematical and engineering), model boundaries (valid operating conditions), etc.

• standardized interfaces (e.g. CAPE-OPEN) to other “liquid/gas simulators”



8

1995-1998 Project „Modellierung komplexer Feststoffprozesse“ (‚Modelling

of complex solids processes‘) funded by VW-Stiftung

→ Development of program and information structures,

programming of simulation system SolidSim,

simulation of exemplary industrial processes,

no graphical user interface

1999-2002 Attempts to get funding of follow-up projects

• Proposal in 5th EU framework

• Proposal for a AiF project (TUHH only)

→ request to submit a proposal for a cooperation project

→ search for project partners

2003 Begin of funding of cooperation project „Fließschema-Simulation

von Feststoffprozessen“ (‚Flow sheet simulation of solids

processes‘) by AiF

• Cooperation of 11 universities

• Advisory committee of industrial partners

Simulation of Solids Processes at TUHH



9

Strategy for Development

• solids processing discipline involves huge number of

different process steps and equipment types

• work load can not be handled by one partner

• solids processing discipline is divided into packages and

distributed for responsible processing to “most

knowledgeable experts” amongst German Universities

• industrial partners of SolidSim project will stand as godfather

for each single package (both, from industrial users & vendors of related equipment)

• throughout project lifetime there will be several field tests of

intermediate results in industry (by industrial partners) to

ensure applicability of system and quality of models



10

SolidSim Developers

COSE GUI, Interfaces G. Gruhn, J. Werther, Hamburg

Size Reduction Mills W. Peukert, Erlangen

Solid-Liquid Separation Filtration

Centrifuge / Decanter

Hydrocyclone

Thickener

S. Ripperger, Dresden

W. Stahl, Karlsruhe

J. Werther, Hamburg

J. Werther, Hamburg

Gas-Solid Separation Gas cyclone

Filter

Scrubber

Electrostatic Precipitator

J. Werther, Hamburg

E. Schmidt, Wuppertal



Classification Sifter

Sieve

Stream classifier


J. Werther, Hamburg

J. Werther, Hamburg

Agglomeration

Granulation

Fluidized Bed Granulator

Pressure Agglomerator

L. Mörl, Magdeburg

Drying Convective Dryer E. Tsotsas, Magdeburg

Crystallisation

Dissolving

Crystallizer M. Kind, Karlsruhe

Conveying

Dossage

dilute phase conveying

dense phase conveying

Dosage

K.-E. Wirth, Erlangen

Spray Spray nozzles P. Walzel, Dortmund

Fluidization Fluidization J. Werther, Hamburg



11

SolidSim Timetable

03

12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3

Selection of test process

Simulation tests with Short-Cut-

Version

Development of detailed models

Model documentation / help

system

Sensitivity analysis

Implementation of methods for

process optimization

Simulation tests using detailed

models

beta testing by selected users

20052004

2. Project Meeting

(09.12.2003)3. Project Meeting

(27.04.2004)

Committal of

detailed models

(30.09.2004)

4. Project Meeting

Committal of

help system

(20.12.2004)



12

SolidSim Simulation Environment

Material Stream Object

Design Concept of SolidSim

• Software consists of

– simulation environment

– material stream objects

– unit models

Unit Model

+ base unit →framework



13SolidSim Simulation Environment


Functions of the SolidSim Framework

• Providing access to model library– allows for easy development of new models by providing standard programming

interfaces and a base module for easy implementation of unit specific models

– dynamic integration of the unit models into the simulation environment

– allows to connect individual models to set up a complex flowsheet

• Graphical user interface– visualization of the flowsheet

– dialog based user interactions

• Simulation engine– coordination of flowsheet calculation

– control of data exchange between unit models and material stream objects during calculation

• Help system– for simple access to documentation of simulation environment and unit models

Unit Model



14SolidSim Simulation Environment


Functions of the SolidSim Framework

• Stream Objects

Transports stream information from one unit model to the next

→ Information Stream: Information about process parameter

→ Energy Stream: Information about transferred energy

→ Material Stream: Information about material stream of physical properties

• Providing access to all stream data necessary for the simulation– Handling of concentrated variables (temperature, pressure) and distributed properties

as well (particle size distribution, etc.)

– Access to physical property package through standardized interfaces

• Conversion and calculation of derived properties– Example: calculation of thermal velocity distribution out of given particle size

distribution and density distribution

Unit Model



15

SolidSim Simulation Environment


Model Library - Guidelines

• models will be constantly maintained and supported by experts from University

• models have to be accurately evaluated and tested in industrial applications

• including godfathers from industry ensures practical applicability and relevance

• models have to have good documentation

• models can be replaced by new improved models easily

• new models can be developed and implemented without access to source code of framework

Unit Model



16• SolidSim utilizes CAPE-OPEN interfaces wherever

possible because of the following benefits:– Software development

• shorter development time using existing interface specification

• Interfaces are validated

– Interoperability• CO compatible units can be used in all COSE‘s

• One COSE might be used as an unit within an other COSE in a flow sheet

• Existing solver, physical property packages, databases can easily be used

• With the existing CAPE-OPEN interfaces one cannot handle distributed solids parameters– TUHH became an associated member of the CO-LaN in February

– proposal for the extension of the standard was submitted by TUHH at the end of 2003

CAPE-OPEN-Standard



17

Folie 3

SolidSim – Structure of Simulation

System

Interface for unit

initialization and

calculation control

Framework: GUI,

model library,

simulation engine

Provides unit

model with

physical

properties and

stream data

Interface for physical

properties and stream

data

Unit specific

simulation

external Physical

Properties System

external Unit Model



18

Inlet

streams

Unit model

Specific model

Base unit

Calculate

Done

Stream data / physical

properties

Simulation Environment

Outlet

streams

ICapeUnit

ICapeThermo+

• Base unit provides functionality

- to simplify communications with stream objects

- to do standard operations (copying inlet to outlet streams)

- ensures CAPE-OPEN 0.9.3, 1.0 and 1.1 compatibility

• Specific unit model

- inherits functionality of the base unit

- implements unit specific simulation only

→ Unit model developers can concentrate on process simulation

SolidSim – Unit Model



19

• global definition of all

compounds used for simulation

• theoretical unlimited number of

phases

– Phase definition

• Phase fraction and

composition

• Definition of discrete

and distributed

properties for each

phase

• MO can utilize any CAPE-OPEN

compatible physical property

package to request physical

property or flash calculations

• Access to global constants


Transports stream information from one unit model to the next and provides accessto physical properties



20

Material Stream Object / Distributed Properties

• Distributed properties are stored in a n-dimensional matrix where n is the number of

defined properties

• Each element of the matrix is defining the fraction of the particles that can be

characterized by a certain combination of properties classes

• The class definition of the distributed properties is stored separately

particle size density form factor

particle size

de

nsity



21

Material Stream Object / Changing

Distributed Properties

• redistribution of mass fractions into existing classes by applying a movement

matrix

• entry in movement matrix represents fraction of the material that is moved from

the old into the new class

• movement matrix is calculated by the unit model and applies only to the properties

the unit requested

• after calculation the unit returns the movement matrix to the framework

• the framework is calculating the new property matrix based on movement matrix

nnn

n

n

kkn

...

kk

kk

n...

,1,

,21,2

,11,1

......

............

......2

......1

21to

from

Example: movement matrix

applied to one distributed

property



22

Extending the Standard

• Extending the existing list of thermo properties

• The functions GetSinglePhaseProp, GetTwoPhaseProp, GetOverallProp will return an interface pointer if a distributed property is requested.

• Adding the following Interface

– ICapeDistributedProperty

• GetDistribution

• GetDistributionState

• GetDistributionClassDefinition

• ApplyMovementMatrix

This interface provides functionality to access distributed properties and to change the fractions by applying a movement matrix.

New classes cannot be added.

The initialization of the distributed properties has to be done through a proprietary interface

orequivalent Set-Functions have to be added to the standard.



23

CapeOpen 1.1 Thermo Interfaces

ICAPEDistributedProperty

+GetDistribution()

+GetDistributionList()

+GetDistributionState()

+GetDistributionClassDefinition()

+ApplyMovementMatrix()

+SetDistribution()

+SetDistributionState()

+SetDistributionClassDefinition()



24

RWK flow sheet

250mm

63mm

100mm

45mm

split size

300 / 225 mmFeed:

Q = 100 t/h

dp50 = 330mm

dp95 = 880mm



25

• Problem:fine solids are highly contaminated

• Solution:Separation of the fines in a hydrocyclone cascade

SolidSim

Application:

sewage sludge Sample:

average contamination

71 ppm

lead, ppm

particle size, mm

mass fraction,

cumulative



26

Hydrocyclone – Separation Efficiency



27

Thank you for your attention!!

Matthias Pogodda

Denickestr. 15

21073 Hamburg

+49 40 42878-3282

The SolidSim Project - CO-LaN– DIVA, gProms, SpeedUp, Aspen Custom Modeler –sequential modular...

Documents

Transcript of The SolidSim Project - CO-LaN– DIVA, gProms, SpeedUp, Aspen Custom Modeler –sequential modular...